24 Nov
2024
24 Nov
'24
9:07 a.m.
On Sunday 24 of November 2024 07:04:02 Andrew Randrianasulu via tde-users
wrote:
вс, 24 нояб. 2024 г., 05:54 Slávek Banko via tde-users
<
users(a)trinitydesktop.org>gt;:
Yes, it is a large volume of code and it can be difficult to get a thorough
overview. But yes, even without obtaining a thorough overview, someone can
do a good finding of a specific bug and prepare a good fix. For example,
thanks to backtrace from the grash. And there is no problem to accept and
merge a good patch from anyone.
Likewise, there is no problem, if someone suggests a patch, we give him
comments that there is a need for some modifications and thanks to the
mutual cooperation between the author and the revising after a while we
move to a good patch, which we then like to merge. This is then beneficial
for both sides - the author gets feedback and will gain better overview
and experience, the project will gain repair and improvements.
However, when someone gives a patch that, for example, removes a line with
a call to release memory, at first glance it looks like an incorrect
solution. When the author on the comment: "This looks like this will
require a more thorough examination to solve the real cause of the
problem," he responds by calling: "No, it works for me, I will not do
anything else! You have to merge it as I did it!”, that's a problem. When,
after some time and thorough examination, it confirms that the problem was
elsewhere and the solution was different, but the author still shouts: "My
fix was correct! You should have merge it as I did it!", this is indeed a
problem to try to cooperate with such an author.
Cheers
Slávek
--
On Sunday 24 of November 2024 00:33:52 deloptes
via tde-users wrote:
There were situations where thorough research clearly showed that the
proposed patch did not solve the cause of the problem, but one
specific consequence - only hides the real cause of the problem. While
the other potential consequences would remain unresolved and would
probably wait for the next hack, which would again solve the
individual consequences, not the cause. While the author fundamentally
refused any cooperation in finding a real cause. As a result, real
repairs of the causes of the problems were merged instead of the
proposed hack.
To be honest I (as non-developer who somewhat forced to be) tend to see
things more from "forever novice" perspective: TDE is big codebase, and
assuming someone can jump right in and prepare professional quality
solution you simply can merge ... is a bit unrealistic?
yes, people might be uncomfortable to extreme if you ask them to do
professional analysis even without telling them *how* it done (assuming
here they, like you, know it by heart).
I guess it hurts both ways .....
I write
nothing to TGW right now since there lie tens my previous
patches!
There are a lot of patches (PRs) in the queue there. No one
complains, except you. It is your free choice. I personally see it
as symbiosis. And at some point of time someone picks up the patch
works it out and it either accepted or rejected.